Auto adjust dynamic exposure and analyze the
key points of face
Face Image Location
Support Facial Features with Different Accuracy
Age, Gender and Hair
Refuse to Recognize Fake Faces
Support the Detection of Safety Helmet
Face recognition products use image processors. It can detect the brightness of the face, automatically adjust the dynamic exposure, track and detect the face, analyze the key points of face, and auto adjust the image magnification, recognition and comparison.
FCARD face recognition technology adopts in-depth learning algorithm, scientifically self-learning and improving algorithm library to improve recognition accuracy, face tracking and detection ability. It obtains the face features through analyze the key points of face, combined with face attributes and other operations to achieve accurate face comparison.
From the in-depth detailed analysis of the eyes, mouth, nose, contour in the face and so on, it can identify 32, 128 and 256 key points. Under the deeply optimized Linux system, it supports the recognition of facial feature points with different accuracy, which can be completed as soon as milliseconds. After optimization, it also can be applied on various scene such as large angle side face, large expression change, partial occlusion, blur, light and shade change, etc.
Our face recognition technology can accurately recognize a variety of face attribute types, such as gender, age, clan, skin color, expression, beard, etc. According to various attribute differentiation, face categories are classified and excluded, so that to speed up the calculation efficiency of face recognition.
We use infrared cameras to detecting infrared inspection on living people, analyze the three-dimensional contour of human faces, and check the three-dimensional changes of human face movements, so that to prevent high-definition photos, screen portraits, portrait masks and other fake fraudulent faces. Realize true and false face recognition.
Sweep the concern Yingze electronic official WeChat